Give us a reason to trust your #INsen poll, Brian Howey
By now, you have probably heard that political columnist Brian Howey has released a poll showing Joe Donnelly ahead of Richard Mourdock by 11 points. But with such a questionable history, why in the world would we trust that poll until we see his sampling data?
Now, please understand that this is not a post to attack Brian Howey. Even if you support the results that Howey shows in one of his polls, EVERY politico in Indiana knows that you must take Howey’s writing and poll results with a grain of salt, lest you look like a complete fool when election day yields the only true poll that matters. And whether it’s Howey or any other source, the devil of polling is always in the details. Here’s why we at HoosierAccess have questions about Howey’s polling sample:
- His poll shows a tie for support of gay marriage (things can always shift, but this is Indiana)
- His poll shows overwhelming support for the decriminalization of marijuana (again, this is Indiana, not California and definitely not one of several more liberal swing states where decriminalization has failed)
- His poll used random dialing instead of calling the actual numbers of registered voters (only registered voters matter on election day)
- His poll called on a Sunday (likely under-sampled conservatives)
- His poll used an unclear method of screening voters (plenty of people have opinions, but you can only count on likely voters to matter on election day)
The bottom line is that the devil is in the details. We need to see the cross tabs and compare them. We don’t know if this poll is like one that you would see on MSNBC, one you’d see on Fox News, or one that was commissioned by a candidate who wants to show off a particular result. Howey led with a story and has not yet backed up his polling. This doesn’t mean that his polling isn’t right, but it does seem questionable when the narrative is so important as we approach election day. It’s extremely important to know if the poll sampled reliable voters, a skewed sample of Democrats, or just called every mouth breather out there.
Howey’s narrative has been clear since before the primary: he supports Dick Lugar and will criticize Richard Mourdock with every chance he gets. It’s not meant as an insult to say so, it’s just an observation of nearly every story that Howey has written about the Indiana Senate race. Yes, Howey did use a pollster who is well-respected for both polls, but pollsters only call when they are told to call and only call the people they are told to call.Let’s not forget that Howey and this same pollster showed Mourdock off by 1o points in the spring.
Questions remain, but with conflicting polls (Howey and Mourdock) it’s difficult to trust anyone right now. What this race comes down it is whether voters think that a Harry Reid controlled Senate, that hasn’t passed a budget in three years and only serves to stonewall economic progress, is good for our country, our economy and national debt.
UPDATE: Howey pollster, Christine Matthews just tweeted the following: “@cmatthewspolls: One thing I would say is #insenpoll shows likely Donnelly win but not by 11 pts on El Day” Pretty interesting admission.
UPDATE 2: Nationally respected pollster Scott Rasmussen has just released a poll that show the race is FAR closer than the sham poll released this morning by Brian Howey. Via Rasmussen.com:
Democratic Congressman Joe Donnelly has a three-point lead over Republican Richard Mourdock in the closing days of Indiana’s U.S. Senate race.
The latest Rasmussen Reports telephone survey of Likely Indiana Voters shows Donnelly with 45% support to Mourdock’s 42%. A surprisingly large number of voters either prefer another candidate in the race (6%) or remain undecided (6%). (To see survey question wording, click here.)
The Indiana survey of 600 Likely Voters was conducted November 1, 2012 by Rasmussen Reports. The margin of sampling error is +/- 4 percentage points with a 95% level of confidence. Field work for all Rasmussen Reports surveys is conducted by Pulse Opinion Research, LLC.